DREADDs in Drosophila: a pharmacogenetic approach for controlling behavior, neuronal signaling, and physiology in the fly

Abstract

We have translated a powerful genetic tool, designer receptors exclusively activated by designer drugs (DREADDs), from mammalian systems to Drosophila melanogaster to selectively, rapidly, reversibly, and dose-dependently control behaviors and physiological processes in the fly. DREADDs are muscarinic acetylcholine G protein-coupled receptors evolved for loss of affinity to acetylcholine and for the ability to be fully activated by an otherwise biologically inert chemical, clozapine-N-oxide. We demonstrate its ability to control a variety of behaviors and processes in larvae and adults, including heart rate, sensory processing, diurnal behavior, learning and memory, and courtship. The advantages of this particular technology include the dose-responsive control of behaviors, the lack of a need for specialized equipment, and the capacity to remotely control signaling in essentially all neuronal and nonneuronal fly tissues.

Figure 1. CNO Has No Effect on Wild-Type Larvae or Adults

(A and B) Mouth hood contractions (A; a measure of feeding), and body wall contractions (B; a measure of locomotion) were counted for 1 min after Canton S (CS) larvae were fed 10% sucrose with no CNO (white bars) or 3.0 mM CNO (black bars). There were no statistical differences in flies fed CNO in either the mouth hook or body wall contractions when compared to control flies (N = 20). (C) Sixteen CS adult males were loaded into the DAMS system with 10% sucrose (white bars) or 10% sucrose with CNO up to 10.0 mM (black bars), and their activity was measured. The activity levels of CNO-fed flies were comparable to control levels. Error bars indicate SEM.

Figure 2. Mammalian DREADD Receptors Couple to Insect G Proteins

(A and B) The hM4Di and rM3BDs receptors were expressed in Drosophila S2 cells in culture, stimulated with CNO, and cAMP levels measured. (A) The percent inhibition of forskolin-stimulated cAMP in S2 cells transfected with the hM4Di receptor. The hM4Di receptor negatively couples to adenylate cyclase, and the hM3BDs receptor positively couples to adenylate cyclase. N = 3 for each data point. (B) Adenylate cyclase accumulation in S2 cells transfected with the rM3BDs receptor.

Figure 3. Activation of Mammalian hM1Dq Receptor Facilitates Neuronal Depolarization

(A–D) The hM1Dq receptor was expressed in flies carrying the elav^c155-GAL4 driver and the UAS-GcAMP. Brains from third-instar larvae were stimulated with KCl (75 mM), either in the presence or absence of CNO (6 µM). GCaMP fluorescence response was measured and the maximum peak latency was determined. (A) In neurons only expressing GCaMP (elav^c155-GAL4 + UAS-GCaMP), pretreatment with CNO had no effect on KCl-induced depolarization. (B) Maximum peak latency was unaffected by pretreatment of CNO in larvae only expressing GCaMP (elav^c155-GAL4 + UAS-GCaMP). (C) In neurons expressing both GCaMP and hM1Dq (elav^c155-GAL4 + UAS-GCaMP + hM1Dq), pretreatment with CNO resulted in a decreased latency to firing. (D) Maximum peak latency was significantly reduced in CNO-pretreated brains expressing hM1Dq (elav^c155-GAL4 + UAS-GCaMP + hM1Dq). *p < 0.005, Student’s t test. N = 15 brains per treatment, ten neurons per brain. Errors bars indicate SEM.

Figure 4. CNO Activation of DREADDs in Chemosensory Neurons Dose-Dependently Controls Behavior

(A–C) Flies carrying either the UAS-rM3BDs, UAS-hM1Dq, or UAS-hM4Di transgene were crossed to flies carrying the SB18.1-GAL4 (6405-GAL4) driver, which drives expression in sensory neurons, and the attraction to ethyl acetate measured. (A) Performance index of flies carrying either the 6405-GAL4 driver, UAS- hM4Di, UAS-hM1Dq, UAS-rM3BDs, or both the 6405-GAL4 driver and the UAS-rM3BDs construct either fed food without CNO (white) or fed food with 1 mM CNO (gray bars). CNO had no effect on any of the parental lines or on flies carrying both 6406-GAL4 and UAS-rM3BDs. (B) CNO activation of hM1Dq in sensory neurons (6405-GAL4 + UAS-hM1Dq) produces a dose-dependent increase in the performance index. (C) Activation of hM4Di in the same neurons (6405-GAL4 + UAS-hM4Di) produces a dose-dependent decrease in the performance index (right). There was no observable difference in the number of immotile animals between all groups. *p < 0.005, ANOVA with Tukey post hoc test for multiple comparison. N = 3, 50 larva per trial. Error bars indicate SEM.

Figure 5. Activation of hM3BDs in Cry Neurons Affects Diurnal Behavior in Adults

(A–C) Males carrying transgenes for the cry-GAL4 driver, the UAS-rM3BDs, or both were fed either 10% sucrose (open circles) or 10% sucrose + 3 mM CNO closed circles) and their activity levels were monitored using the DAMS system for 5 days. (A and B) Flies carrying either the UAS-rM3BDs (A) or the cry-GAL4 (B) exhibited no significant alterations in diurnal behaviors when fed CNO as compared to the flies not fed CNO. (C) F1 flies expressing rM3BDs in cry-GAL4 neurons (cry-GAL4 + UAS-rM3BDs) fed CNO exhibited decreased late day activity levels compared to non-CNO-fed F1 flies. Gray shading indicates dark conditions (ZT 0 = lights on). Results represent the average of two (A and B) and three (C) separate experiments, with each separate experiment consisting of 16 male adult flies each for each treatment condition. *p < 0.05, two-way ANOVA with Bonferroni post hoc test for multiple comparison. Error bars indicate SEM. There appears to be an unknown mutation/background effect affecting normal diurnal behavior in the parental cry-GAL4 driver. Nevertheless, the F1 progeny demonstrate normal diurnal behavior in the absence of CNO, which is affected by CNO, whereas the parental behavior is not.

Figure 6. Activation of hM4Di in the Mushroom Bodies Disrupts Short-Term Learning and Memory

(A) Flies carrying either the UAS-hM4Di transgene (left), the MB247-GAL4 driver (center), or both in combination (right) were fed CNO and the STM performance measured [PI = (# of flies avoiding the paired odor) – (# of flies avoiding the unpaired odor)/total flies]. The parental strain flies carrying only one or the other of the transgenes exhibited no difference in the performance index after being fed CNO. The F1 flies carrying both transgenes exhibited a 50% decrease in the performance index when the expressed hM4Di receptor (MB247-GAL4 + UAS-hM4Di) was activated with 1 mM CNO. *p < 0.05, Student’s t test. N = 8 trials. Error bars indicate SEM. (B) The effects of DREADD activation are reversible. Flies expressing hM4Di in the mushroom bodies (MB247-GAL4 + UAS-hM4Di) were again monitored for short-term memory, this time after being removed from the CNO. Flies immediately trained and tested exhibited the expected 50% decrease in performance index (dark gray). When flies were removed from the CNO for 3 hr, the decrease in PI was only 12% (gray). Flies that were removed from the CNO for 12 hr prior to training and testing did not exhibit a decrease in PI (light gray) when compared to F1 flies that were not fed CNO (white). *p < 0.05, ANOVA with Bonferroni post hoc test for multiple comparison. N = 8. Error bars indicate SEM.

Figure 7. hM4Di Activation in 5-HT₇-Dro Circuitry Dose-Dependently Disrupts Courtship and Mating

(A–C) Flies expressing the hM4Di receptor under the control of the 5-HT₇Dro-GAL4 driver were fed CNO and the effects on courtship index tested. (A and B) CNO (1.0 mM in the food for the 48 hr prior to testing) had no effect on either parental strain. (C) Flies expressing the hM4Di receptor (5-HT⁷Dro-GAL4 + UAS-hM4Di) and fed CNO in the food at the indicated concentrations for 48 hr prior to testing had their courtship index dose-dependently disrupted, consistent with our previous results demonstrating a role of this neuronal circuit in courtship and mating behaviors. Courtship index was calculated by the amount of time a pair of flies spent engaged in courtship rituals divided by the total amount of time of the assay (10 min), expressed as a percentage. *p < 0.05, ANOVA with Bonferroni post hoc test. N = 10–20 mating pairs per treatment. Error bars indicate SEM.